There is a plethora of data available to users on the Internet today. With so much information, the user is faced with an information overload. A need exists for efficient filtering mechanisms that give better and faster access to data. Some sites (e.g., CNN and Quicken) address this problem by allowing a user to create personalized pages that contain only information that is of interest to the user. Other sites send notifications when the underlying data changes and some condition is met. Some problems with these approaches to personalization include: 1) personalization is limited to the information accessible from a single site; 2) personalization features are not offered by all Web sites that might be of interest to a user; and 3) since sites that provide personalization require users to sign up, privacy issues may be involved because the user needs to divulge personal information when signing up thereby potentially enabling the sites to track the interests of the user.